The Pioneer does send a team to the playoffs. Perhaps your analysis is suggesting (might suggest) this should be a more-than-one-bid league. That is why they should not be "removed". Ivies should also not be removed, because this provides a comparison point, especially for the few teams that may have played an Ivy and might be in playoff consideration.
My point here is that as a scientist, we deal with data sets and there are outliers that we maybe should (or should not) ignore. For example, and there is no political agenda here (so please don't turn it into one), the COVID vaccine has a measured rate of inflammatory heart disease of 2.3 in 100,000 people vaccinated (
https://pmc.ncbi.nlm.nih.gov/articles/PMC8755376/). Just like COVID infection has a measured rate of heart inflammation approximately 7 times that of those vaccinated (
https://www.frontiersin.org/journals...22.951314/full). That said, that is ~15 in 100,000 people. My point: significance is in the eye of the beholder, and just because less than 0.002% show heart inflammation after vaccination, does not mean that it does not matter. Just like one could also say that 0.002% is much lower than the mortality rate for COVID, which some might also argue is low enough to matter or not matter. It is simply perspective and interpretation of data. Applying this to football, that is why considering only the top 7.75% of teams in any category is perhaps not the most rigorous way to consider the data.
Got a kick out of this! I think this might be able to be done via a Google Sheet (I do this with a few columns of data for other things, but would have to play around with 47 columns, or whatever you said were the number of stats).
Cherry-picking was perhaps not the best way to describe it; however, without including ALL data, it will always have a level of bias that including all data would not. If you add in "weighting", that would also introduce bias. Heck, then one might as well use SOS.
The objective data has yet to determine this; however, since I do believe that strength of opponents (and thus schedule) differs interconference and can even have a moderate-to-great disparity intraconference, some teams might make higher appearances in statistical items, such that one might have at least a little WTF response. Is it possible that for NDSU, SDSU, MTST, USD, UCD, etc, that while consensus is that they are at the top, essentially averaging out all the statistical categories might not necessarily reflect that...and it might even favor those good teams in a perhaps not so good conference (but without SOS, there would be no such thing as a not so good conference, right?). At this point, it is speculative, nothing more.
I appreciate that. Top 10 was an arbitrary cutoff, but meaningful? This remains to be seen. I do agree that there appears to be at least a little more than coincidence to this observation.
Bookmarks